NL2016301B1

NL2016301B1 - Method for detecting a marker for active tuberculosis.

Info

Publication number: NL2016301B1
Application number: NL2016301A
Authority: NL
Inventors: Theodorus Castrop Johannes
Original assignee: Kei Int Ltd; Tomorrows Ip Ltd
Priority date: 2016-02-22
Filing date: 2016-02-22
Publication date: 2017-08-25
Also published as: WO2017143985A1

Abstract

The present invention relates to a method of detecting a marker for active tuberculosis, and a method for pre-treating a sample from a human or animal suspected of having active tuberculosis. The present invention aims to overcome the problems that derive from the binding of materials that are not indicative for active tuberculosis to immobilised mycolic acid antigens.

Description

Method for detecting a marker for active tuberculosis

The present invention relates to a method of detecting a marker for active tuberculosis, and a method for pre-treating a sample from a human or animal suspected of having active tuberculosis.

Introduction

Mycobacterium tuberculosis is a pathogenic bacterial species in the family Mycobacteriaceae and the causative agent of most cases of tuberculosis (TB).

Nine million people fell ill with TB in 2013, including 1.5 million cases among people with HIV. In 2013, 1.5 million people died from TB, including 360000 among people who were HIV-positive. TB is one of the top three killers of women worldwide, 510000 women died from TB in 2013. Of the TB deaths among HIV-positive people, 50% were among women. At least 550000 children became ill with TB and an estimated 80 000 children who were HIVnegative died of TB in 2013. Globally in 2013, an estimated 480000 people developed multidrug-resistant TB (MDR-TB) and there were an estimated 210000 deaths from MDR-TB. At least one case of extensively drug-resistant TB (XDR-TB) has been reported by 100 countries by the end of 2013. On average, an estimated 9% of MDR-TB cases have XDR-TB. A reliable and fast way of diagnosing tuberculosis is therefore of utmost importance.

Several methods of diagnosing tuberculosis have been developed, but all methods have their disadvantages.

Diagnosing active tuberculosis based merely on signs and symptoms is difficult, as is diagnosing the disease in those who are immunosuppressed. The TB skin test (also called the Mantoux tuberculin skin test), TB blood tests (also called interferon-gamma release assays or IGRAs), and chest radiography (X-ray), and tests on the presence of acid-fast-bacilli (AFB) on a sputum smear, indicate some of the infected individuals in days. A definitive diagnosis of TB is made by identifying M. tuberculosis in a clinical sample (e.g. sputum, pus, or a tissue biopsy). However, the difficult culture process for this slow-growing organism can take two to six weeks for blood or sputum culture.

Humans or animals infected with the M. tuberculosis normally produce antibodies directed against the Mycobacterium. The presence of these antibodies in a sample taken from infected individuals indicates the infection. For instance, WO 2005/116654 and WO 2013/186679 describe methods based on this principle and disclose methods of detecting antigen specific biomarker antibodies for the diagnosis of active tuberculosis, by detecting binding of antibodies against mycolic acid antigens to immobilised mycolic acid antigens.

The inventor has observed that when a method of detecting a marker for tuberculosis, which is based on binding of antibodies against mycolic acid antigens in a sample to immobilised mycolic acid antigens, is applied to a sample derived from a healthy subject, there is also a high degree of binding of materials to immobilised mycolic acid antigens. A sample from a healthy subject should not contain antibodies directed against Mycobacteria, such as antibodies against mycolic acid antigens. This high degree of binding to immobilised mycolic acid antigens of materials present in the samples of healthy subjects thus indicates that samples derived from healthy subjects contain materials which bind to mycolic acid antigens, but which are not indicative for tuberculosis. In a test of detecting markers for active tuberculosis this will inevitably lead to a high risk of false positive results so that a healthy subject may incorrectly test positive for active tuberculosis. In case a subject would test positive such unreliability would necessitate further follow up tests to confirm whether the subject indeed suffers from tuberculosis, before starting lengthy, expensive and often unhealthy treatment with antibiotics against Mycobacteria.

In case a subject actually suffers from tuberculosis, the further problem arises that if materials that are not indicative for active tuberculosis bind to immobilised mycolic acid antigens in a detection test a high background binding signal is produced. This high background signal obscures the signal derived from the actual markers for tuberculosis.

The present invention aims to overcome the problems that derive from the binding of materials that are not indicative for active tuberculosis to immobilised mycolic acid antigens.

Summary of the invention

The aim of the invention has been achieved by the provision of the methods of the invention.

In a first aspect the invention relates to a method of detecting a marker for active tuberculosis, comprising the steps of: I) providing a sample from a human or animal suspected of having active tuberculosis; II) filtering the sample using a cut-off membrane filter with a nominal pore size of 10 nm or more and less than 100 nm; III) exposing at least part of the filtered sample to a substrate carrying an immobilised mycolic acid derived antigen; and IV) detecting binding of antibodies to said mycolic acid derived antigen.

In a second aspect the invention relates to a method for pre-treating a sample from a human or animal suspected of having active tuberculosis for detection, comprising the steps of: I) providing a sample from a human or animal suspected of having active tuberculosis; and II) filtering the sample using a cut-off membrane filter with a nominal pore size of 10 nm or more and less than 100 nm.

The inventor has surprisingly found that if a sample from a human or animal is filtered using a cut-off membrane filter with a nominal pore size of 10 nm or more and less than 100 nm before being subjected to a detection substrate carrying an immobilised mycolic acid derived antigen and detecting binding of antibodies to the antigen, the binding signal of materials that are not indicative for active tuberculosis to immobilised mycolic acid antigens is significantly reduced. This reduces the risk of that healthy subjects test positive for active tuberculosis. Furthermore because applying a filter step to the sample using a cut-off membrane filter with a nominal pore size of 10 nm or more and less than 100 nm filters results in a significant reduction of materials that are not indicative for active tuberculosis, the background signal derived from these materials is markedly reduced. The signal derived from the actual markers for tuberculosis is therewith significantly less obscured by the background signal, so that the signal derived from the actual markers for tuberculosis becomes more pronounced.

Description of the figures

Fig. 1 shows a diagram with results of tests with cut-off membrane filter with different cut-off sizes and samples derived from healthy persons (Control) and persons having active tuberculosis (TB+).

Detailed description

The invention relates to a method of detecting a marker for active tuberculosis, comprising the steps of: I) providing a sample from a human or animal suspected of having active tuberculosis; II) filtering the sample using a cut-off membrane filter with a nominal pore size of 10 nm or more and less than 100 nm; III) exposing at least part of the filtered sample to a substrate carrying an immobilised mycolic acid derived antigen; and IV) detecting binding of antibodies to the antigen. The invention further relates to a method for pre-treating a sample from a human or animal suspected of having active tuberculosis so that the sample can be used in a detection assay in which antibodies against mycolic acid derived antigens are detected, comprising the steps of: I) providing a sample from a human or animal suspected of having active tuberculosis; and II)filtering the sample using a cut-off membrane filter with a nominal pore size of 10 nm or more and less than 100 nm.

For purposes of diagnosis of tuberculosis, one or more samples from a human or animal suspected of having active tuberculosis may be compared to a sample from a human or animal which is confirmed to be healthy. For the sake of reliability it is highly preferred that all samples in one analysis undergo the same treatment in accordance with the steps of the method of the invention. The strong decrease in background signal and the associated lower risk of false positives make it even possible to reliably diagnose whether a person has tuberculosis without the necessity of a reference sample obtained from a healthy subject. The strong decrease in background signal and the associated lower risk of false positives also make it possible to reliably diagnose whether a person has tuberculosis without the necessity to divide a sample from a human or animal suspected of having active tuberculosis into two sample fractions of which one is exposed to mycolic acid antigens before the two fractions are exposed to a detection substrate with mycolic acid antigens such as for instance described in WO 2005/116654 or WO 2013/186679.

In the methods of the invention a sample from a human or animal suspected of having active tuberculosis is provided. These samples may in particular be blood samples of a human or animal suspected of having active tuberculosis. The sample is preferably a whole blood sample. The sample may be obtained by any regular means of obtaining blood from a subject. In order to be used in the methods of the invention samples may be used that have been collected at an earlier stage, stored until use under suitable conditions and provided at a suitable moment. Alternatively, a sample may be used in the detection method of the invention on the spot, i.e. as a point of care test.

In case the sample is a whole blood sample, the sample is preferably pre-filtered or separated to plasma or serum before the step of filtering the sample using a cut-off membrane filter with a nominal pore size of 10 nm or more and less than 100 nm. A suitable filter for such a pre-filtering step is a 0,2 micron filter. Pre-filtering the sample with a filter with larger pores than the cutoff membrane filter also prevents that aggregates of materials such as blood cells cause clogging of the cutoff membrane filter.

The sample is preferably a blood derived sample. The sample may be a whole blood sample, a plasma sample or a serum sample. Blood serum is blood plasma without clotting factors and is preferred as plasma. The word plasma in this application may therefore as well refer to (blood) serum.. Serum is preferred because it contains less different materials than blood plasma which may lead to aspecific interactions or unwanted biological activity. In addition serum may have a lower viscosity than blood plasma. Using serum therefore may circumvent the need for diluting a sample, which saves time and materials.

About 55% of whole blood consists of plasma/serum. If a whole blood sample is not pre-filtered perfectly or if the patient's physical situation necessitates it, it may be desired to dilute the whole blood sample or plasma or serum. The words plasma or serum in this application may therefore also refer to diluted plasma or serum. A dilution of the blood or plasma may therefore be implemented in the method of the invention, such as 5 to 10 x dilution, a 10 to 20 x dilution, a 20 to 50 x dilution, a 50 to 100 x dilution, a 250 to 5000 x dilution, a 750 to 1250 x dilution, such as for instance a 5 x, 10 x, 20 x, 50 x, 100 x, 200 x, 500 x, 4000 x, 2000 x or 1000 x dilution. Depending on the viscosity of the sample, such dilution may take place before the step of separating the plasma from the blood step (pre-filter step or separating step). Alternatively dilution may take place after the pre-filter step or separating step, and before the step of filtering the sample using a cut-off membrane filter with a nominal pore size of 10 nm or more and less than 100 nm. The dilution may also be performed as an alternative to said pre-filter/separating step, and before the step of filtering the sample using a cut-off membrane filter with a nominal pore size of 10 nm or more and less than 100 nm.

Dilution may be performed with any suitable diluent, for example a PBS based buffer, such as a blocking buffer. Such buffer may for example be a PBS/AE buffer comprising NaCl, KC1, KH2P04, Na2HP04 and EDTA in water at physiological pH. Such buffer may be a PBS based buffer consisting of 8.0 g NaCl, 0.2 g KC1, 0.2 g KH2P04, and 1.05 g Na2HP04 per liter of double distilled, deionized water containing 1 mM EDTA and 0.025% (m/v) sodium azide which is adjusted to pH 7.4.

The whole blood sample or plasma or serum may be further diluted with agents that prevent blood clotting, such as EDTA, heparine or citrate.

Optionally a detergent may be added in low concentration to the blood/plasma/serum to avoid sticking of components of the test system used.

In the step of filtering the sample using a cut-off membrane filter with a nominal pore size of 10 nm or more and less than 100 nm is used. The inventor has observed that it is this range of pore sizes that results in a significant reduction of materials that are not indicative for active tuberculosis in a sample, while antibodies that are indicative for tuberculosis are not reduced or reduced to a lesser extent so that the background signal derived from these materials is markedly reduced. In a sample derived from a subject with active tuberculosis the signal derived from the actual markers for tuberculosis is therewith significantly less obscured by high background signal, so that the signal derived from the actual markers for tuberculosis becomes more pronounced.

By nominal pore size is meant average pore size, in which more than about 90% of the pore sizes are within the range of about plus or minus 10% of the nominal pore size, e.g., such as plus or minus 1 nm in the case of 10 nm nominal pore size. Nominal pore size as referred to in this application are sizes as determined by electron microscopy.

It is preferred that the cut-off membrane filter has a nominal pore size between 25 and 75 nm, such as between 25 and 70, 65, 60, 55 or 50 nm. Even more preferred is that the nominal pore size is between 25 and 45 nm, such as between 30 and 40 nm, between 33 and 38 nm or between 34 and 36 nm. It is even more preferred that the nominal pore size is about 35 nm. In these preferred ranges the amount of materials that are not indicative for active tuberculosis in a sample are highly reduced, whilst still maintaining a sufficiently high amount of actual markers for tuberculosis in the filtrate. It is important that the nominal pore size of the membrane is such, that at least a measurable extent of IgG antibodies against mycolic acid derivatives is able to pass to through the membrane. If a pore size of less than 10 nm is used the amount of materials that are not indicative for active tuberculosis in a sample would be significantly reduced, but the amount of actual markers for tuberculosis has also been scavenged away to such an extent that the filtrate becomes unsuitable for detection of markers for active tuberculosis .

The membrane of the cut-off membrane filter is preferably a hydrophilic membrane. The inventor has observed that this leads to reliable results. Suitable hydrophilic materials for the membrane of the cut-off membrane filter may be polyethersulphone, polyvinylidene difluoride and regenerated cellulose. These materials are low protein binders so that antibodies against mycolic acid antigens do not bind to the membrane. Polyethersulphone membranes are preferred because these provide higher flow rates and lower protein binding than competitive regenerated cellulose. This results in lower processing times and the highest possible recoveries of antibodies against mycolic acid antigens which are specific for active tuberculosis in the filtered sample.

Suitable cut-off filters can be obtained commercially. For instance cut-off filters with Molecular Weight Cut Off (MWCO) of 100 KDa, 300 KDa and 1000 KDa are available from Pall Corporation (Nanosep®). These MWCOs correspond to membrane nominal pore sizes of 10 nm, 35 nm and 100 nm respectively.

Filtering the sample using abovementioned cut-off membrane filter can take place by means of any suitable filter method, using any suitable filter device. Filtering can take place by gravity or capillary force, but is preferred that filtering takes place using ultracentrifigation, such as centrifugation at 10.000 g. This way filtering takes place with few seconds, thus reducing the time of treatment of the sample, so that diagnosis of tuberculosis can take place in a fast way.

In the cut-off membranes of the methods of the invention substances that are larger than a specified filter "cut-off" are retained by the filter and smaller substances pass through the filter with the filtrate.

After filtering the sample using a cut-off membrane filter with a nominal pore size of about 10 nm or more and less than 100 nm, the filtered sample or part of it can be used directly for detection of antibodies against mycolic acid antigens or be stored until further use. With filtered sample is meant filtrate, i.e. the part of the sample that flows through the membrane.

It is preferred that at least part of the sample is exposed to a sterol lipid, prior to exposing it to the detection substrate carrying an immobilised mycolic acid derived antigen. Preferably at least part of the sample is exposed to said sterol lipid after filtering it with said cut-off membrane filter, and prior to exposing it to said substrate carrying an immobilised mycolic acid derived antigen. The inventor has found that this results in even more reduction of materials that are not indicative for active tuberculosis in a sample, while the background signal derived from these materials is even more reduced, thus lowering the risk of false positive tuberculosis detection even more.

The exposure time of the sample to the sterol lipid is preferably less than 10 minutes, such as 2 to 8, 3 to 7, 4 to 6 or about 5 minutes. The exposure time depends on the way the sample is brought into contact with the sterol lipid.

The sterol lipid used in the context of the invention preferably is cholesterol or a derivative thereof. The sterol lipid may also be a sterol modified phospholipid. Such sterol-modified lipid may a sterol-modified phospholipid, for instance a sterol-modified phosphatidylcholine lipid or glycerophospholipid. In such sterol modified lipid the sterol is preferably cholesterol. A good example of a sterol-modified lipid suitable for the purposes of the invention is 1-palmitoyl-2-cholesteryl carbonoyl-sn-glycero-3-phosphocholine.

The sterol lipid is preferably immobilized on a surface. An example is a substrate having a coating containing cholesterol or cholesterol ester wherein the cholesterol ester is cholesterol linoleate, wherein a weight ratio of linoleic acid to cholesterol is in the range from 1:3 to 1:20. One exemplary way of exposing the sample to the sterol lipid is to expose the sample to beads that are pre-coated with a sterol lipid, which is preferably cholesterol. A substrate may also be coated with a sterol lipid, preferably cholesterol, in combination with other molecules .

Preferably said sterol lipid is cholesterol immobilized on a substrate together with phosphatidyl choline. The sterol lipid scavenges away the anticholesterol antibodies from the blood/plasma/serum that would otherwise cross react with the mycolic acid antigens on the substrate and lead to false positive diagnosis of tuberculosis. Phosphatidyl choline will bind to hydrophobic materials in the blood sample, rendering the sample more hydrophilic after exposure. The resulting hydrophilic sample will easier to be handled in the subsequent method steps and be less prone to clotting.

Sterol lipids, preferably cholesterol may be immobilised together with pectin on a substrate, e.g. the inner wall of a tubing or on beads. In this embodiment the sterol lipid scavenges away the anti-cholesterol antibodies from the blood/plasma/serum that would otherwise cross react with the mycolic acid antigens on the substrate and lead to false positive diagnosis of tuberculosis. In addition pectin scavenges away cholesterol in the sample and therewith also cholesterol antibodies. This results in even more reduction of materials that are not indicative for active tuberculosis in a sample, while the background signal derived from these materials is even more reduced, thus lowering the risk of false positive tuberculosis detection even more.

Alternatively a sterol lipid, preferably cholesterol is immobilised together with a compound binding to cholesterol in the blood derived sample, such as cholesterol binding heteropolysaccharide, such as β-cyclodextrin, pectin, amphotericin B or dextrin. Such molecules scavenge away cholesterol in the sample and therewith also cholesterol antibodies. This results in even more reduction of materials that are not indicative for active tuberculosis in a sample, while the background signal derived from these materials is even more reduced, thus lowering the risk of false positive tuberculosis detection even more.

For instance, a sterol lipid, preferably cholesterol is immobilised with β-cyclodextrin or pectin, amphotericin B on a substrate, such as hollow fiber polypropylene membranes or glass or beads.

Many methods are known to immobilise a sterol lipid, in particularly cholesterol on a substrate. The skilled person will be able to select the protocol suitable for his particular coating.

For detection at least part of the filtered sample is exposed to a substrate carrying an immobilised mycolic acid derived antigen, i.e. to a detection substrate, and binding of antibodies to the antigen is detected.

Detection of binding of antibodies to the antigen on the detection substrate may take place in real time or by means of an end-point assay.

Suitable real time detection assays include surface plasmon resonance or electrochemical impedance spectroscopy, isothermal titration calorimetry, bio-layer interferometry, optical gratings, photonic crystal, acoustic resonant profiling, quartz crystal microbalances.

In a real time detection method, preferably the substrate carrying an immobilised mycolic acid derived antigen which is used for detecting binding of antibodies to the antigen is silica based, such as substrates based on silicium dioxide. Silica based substrates are particularly useful when ring resonance technology is used to detect binding of antibodies to the immobilised mycolic acid antigens. Preferably the detection is carried out using a biosensor chip using Si-based ring resonator. This enables the method of the invention to be carried out with a very compact device. It is also well possible that the substrate carrying an immobilised mycolic acid derived antigen which is used for detecting binding of antibodies to the antigen is gold based. Gold based substrates are particularly useful when surface plasmon resonance or electrochemical impedance spectroscopy are used to detect binding of antibodies to the immobilised mycolic acid antigens .

In a real time method, because detection takes place in real time, binding of antibodies on the detection substrates is directly detected during the binding process. For detection in principle all real-time, label free analysis techniques may be used.

The detection of binding of antibodies and/or other material to the mycolic acid antigen on the detection substrate may be carried out in an automated device. Various automated devices will be known to the person skilled in the art and the skilled person will be able to select suitable software means to determine the degree or extent of binding to the detection substrate.

In an exemplary embodiment the method of the invention, when using a real time detection method, comprises i) providing a sample from a human or animal suspected of having active tuberculosis; ii) filtering the sample using a cut-off membrane filter with a nominal pore size of about 10 nm or more and less than 100 nm; iii) exposing the filtered sample to a sterol lipid; iv) obtaining at least two fractions of said sample either before or after exposing it to said sterol lipid; v) exposing the first of said fractions to a substrate carrying an immobilised mycolic acid derived antigen; vi) exposing the second of said fractions to a substrate not carrying an immobilised mycolic acid derived antigen; vii) exposing the sample fraction exposed in step v) to a test substrate carrying an immobilised mycolic acid derived antigen and exposing the sample fraction exposed in step vi) to a control substrate carrying an immobilised mycolic acid derived antigen; viii) detecting binding of antibodies to the antigen of step vii) in real time; and ix) comparing the degree or extent of binding between the test and control substrates, any observed lesser binding to the test substrate being an indicator of the presence of antibodies to the antigen in the sample that indicates active tuberculosis in the human or animal from which the samples originated. In this context it should be understood that lesser can be interpreted qualitatively and quantitatively, i.e. lesser binding may be interpreted as having less binding events as well as having weaker bindings. The advantage of this embodiment is that no sample of a healthy person is required as a reference or control sample, even if there is still background signal.

Only one sample from one subject is necessary in this embodiment. The test detection substrate and the control detection substrate in this embodiment may be separate entities. Alternatively the test substrate and the control substrate may be realized as different positions (spots) on one substrate entity. The sample from a human or animal suspected of having active tuberculosis of step i) of this embodiment is as described above. One exemplary way of exposing the sample to the sterol lipid in this embodiment is to lead the sample into a long spiral channel and pass it along the length of the channel, which is pre-coated with a sterol lipid, which is preferably cholesterol. At the end of the channel the lipid sterol exposed sample may pass a means for dividing the sample stream such as a passive valved branch point that leads the divided sample streams to the next substrate carrying an immobilised mycolic acid derived antigen of step vi) and the substrate not carrying an immobilised mycolic acid derived antigen of step vii), e.g. a container comprising a substrate carrying an immobilised mycolic acid derived antigen, and a container comprising a substrate not carrying an immobilised mycolic acid derived antigen respectively. An alternative exemplary way of exposing the sample to the sterol lipid is to inject it into a container coated with cholesterol, followed by incubation therein for less than 10 minutes. Whole blood/plasma/serum is rich in hundreds different kind of molecules with hydrophilic to hydrophobic properties. Therefore substrate material will be used that is inert for non-specific binding of molecules of this sample. In this context the substrate should be understood to be a support material. In the steps v) and vi) of this embodiment a fraction of the sterol lipid exposed sample is exposed to a substrate carrying an immobilised mycolic acid derived antigen (step v) ) and another fraction of the sterol lipid exposed sample to a substrate not carrying an immobilised mycolic acid derived antigen (step vi)) . To obtain at least two fractions, at a certain moment in the process the blood/plasma stream has to be divided (step iv of this embodiment). For sake of convenience and to optimize reliability of the method, it is preferred that the stream is divided after exposure to the sterol lipid, i.e. that the sterol lipid exposed sample is divided into at least two fractions to provide a test stream and a control stream. It is preferred that the sample is divided in two equal or substantially equal fractions, because this enables a simple comparison of both fractions without the need for correction calculations. Dividing the sample stream may be carried out by any means for dividing the sample stream such as a passive valved branch point. Alternatively the sample stream (which is pre-filtered to plasma or serum using a filter with larger pore size than 100 nm as described above, such as a microfilter of 0,2 micron) is divided into at least two fractions after the pre-filtering step, but before exposure to the sterol lipid. Alternatively the blood stream is divided before pre-filtering. This is less preferred because in this case two microfilters would be required. The substrate carrying an immobilised mycolic acid derived antigen and the substrate not carrying an immobilised mycolic acid derived antigen of steps v) and vi) are preferably of the same material. In addition, since sterol lipids, (e.g. cholesterol) and mycolic acid derivatives are both hydrophobic, the substrates for exposure in steps v) and vi) will be of the same material as the substrate that may be used in step iii). It will be understood that suitable substrate material is inert for non-specific binding of molecules of this sample. One exemplary way of exposing the sample to the substrates of step v) and vi) of this embodiment is to lead the divided sample streams into long spiral channels, preferably implemented in micro-chips, pre-coated either with a mycolic acid derivative (step v) or without a mycolic acid derivative (step vi) and pass it along the length of these channels. The distance of both channels to the biosensor with the detection substrates will need to be of the same length. It is important that there is as less as possible non-specific binding the substrate material. An alternative exemplary way of exposing the divided lipid sterol exposed samples to a substrate coated either with a mycolic acid derivative or without a mycolic acid derivative is to inject the a first stream into a container comprising a substrate carrying an immobilised mycolic acid derived antigen (step v) , and a second sample stream into a container comprising a second substrate not carrying an immobilised mycolic acid derived antigen (step vi) and incubate the samples for less than 10 minutes, such as 2 to 8, 3 to 7, 4 to 6 or about 5 minutes .

Suitable end point assays include enzyme-linked immunosorbent assay (ELISA) , Western blotting, radioactive labelling assay, photospectrometric assay, immunofluorescence, immunoprecipitation, immunocytochemistry, immunohistochemistry, electrochemical impedance spectroscopy.

In a preferred embodiment of an end point assay detection takes place by means of an immunogold filtration assay. In such an assay the detection substrate is a microporous membrane, preferably a nitrocellulose membrane or a PVDF membrane, which is coated with an immobilised mycolic acid derived antigen.

In an end-point assay interaction of antibodies with the mycolic acid derived antigens may be carried out using secondary antibodies that bind the heavy chain of the antibodies against the mycolic acid derivatives. Many suitable secondary antibodies are commercially available. The secondary antibody may be coupled to beads, for instance gold beads, or associated with liposomes. Examples of secondary antibodies may be protein A or G, possibly conjugated with an enzyme that enables detection. A particular suitable technique or detecting the binding of antibodies to the immobilised mycolic acid antigens on the detection substrate in an end-point assay is the so-called immunogold filtration assay (IGFA), and in particular the dot immunogold filtration assay (DIGFA).

Immunogold filtration assays are methods combining ELISA and immunogold technique and are methods in which a sample to be assayed is allowed to filtrate through a microporous membrane, preferably a nitrocellulose membrane, and is captured by a capture probe coated on the membrane. A colloidal gold labelled probe is allowed to filtrate through the microporous membrane in the same manner. By using a microporous membrane as the carrier for the capture probe and employing the capillary action and permeability of the membrane antigens and antibodies can easily react and may conveniently be subjected to optional washing and/or blocking steps. When the colloidal gold labelled probe binds to the capture probe the colloidal gold particles aggregate and a red dot appears which is visible with the naked eye.

Immunogold filtration assays are simple and rapid detection methods because no instruments are required except a membrane and the reagents and the results can be observed by the naked eye within a few minutes.

In an immunogold filtration assay the microporous membrane may be for example a nitrocellulose membrane, a cellulose acetate membrane or a PVDF membrane with a suitable pore diameter. Preferably nitrocellulose is used. A suitable pore diameter is 0,2 to 5 pm.

In case the end-point assay is an immunogold filtration assay, the detection substrate is a microporous membrane, preferably a nitrocellulose membrane, which is coated with a mycolic acid derived antigen. After immobilizing the mycolic acid derived antigen onto the microporous membrane, the pretreated samples can be applied to the membrane. After addition of the sample fractions and reaction of the immobilized mycolic acid derived antigens with the antibodies contained in the samples on the membrane, colloidal gold-labeled second antibodies can be added onto the membrane to have gold particle aggregation in the antigen-antibody reaction place. In case of aggregation visible red or brown spots are formed. The intensity of the spot is proportional to the amount of reactions between antigen and antibody, i.e. to the amount of antibodies in the pre-treated sample. In other words a sample from a person suffering from tuberculosis will result in a more intense spot than a sample from a person which is healthy.

Between the various steps of an immunogold filtration assay the membrane may be washed with a suitable buffer, for example a PBS based buffer. Such buffer may for example be a PBS/AE buffer comprising NaCl, KC1, KH2P04, Na2HP04 and EDTA in water at physiological pH. Such buffer may be a PBS based buffer consisting of 8.0 g NaCl, 0.2 g KC1, 0.2 g KH2P04, and 1.05 g Na2HP04 per liter of double distilled, deionized water containing 1 mM EDTA and 0.025% (m/v) sodium azide which is adjusted to pH 7.4.

In case DIGFA is used, the microporous membrane may be coated with said mycolic acid derived antigen in a dot wise manner. In a DIGFA assay the pre-treated samples are applied to the membrane in the form of dots. Also the colloidal gold-labelled second antibodies are added in the form of dots. A DIGFA assay is particularly preferred because at different spots on several membranes various antigens deriving from various mycobacterial strains may be immobilized. This way it becomes possible to provide information on which mycobacterial strain a patient is infected with. Another advantage of using DIGFA is that samples derived from different persons suspected of having active tuberculosis can be compared in one test, because DIGFA enables fast and reliable detection of antibody-antigen interaction in an unlimited amount of spots, depending on the size of the membrane.

The detection of binding of antibodies and/or other material to the mycolic acid antigen, for instance the red staining in case a DIGFA assay is used as a detection method, may be carried out in an automated device. Various automated devices will be known to the person skilled in the art and the skilled person will be able to select suitable software means to quantify the degree or extent of binding on the detection substrates.

The detection of binding of antibodies and/or other material to the mycolic acid antigen may be performed by a visual detection technique or any other suitable detection technique. In a particular preferred embodiment, detection by means of the end-point assay takes place visually, preferably with the naked eye. This has the advantage of easy detection without the need for expensive and complicated detection technology. In case DIGFA is used binding of antibody antibodies and/or other material to the mycolic acid antigen may be assessed by means of the naked eye. A visual signal, e.g. the red staining in case a DIGFA assay is applied as end-point assay, may also be detected with help of a mobile app, i.e. a computer program designed to run on mobile devices such as tablet computers or smart phones. For instance, an app can be used that is designed to compare the binding signal between different samples or sample fractions and which indicates whether the human or animal from which the sample originated has active tuberculosis.

In an exemplary embodiment the method of the invention, when using an end-point detection method, comprises i) providing a sample from a human or animal suspected of having active tuberculosis; ii) filtering the sample using a cut-off membrane filter with a nominal pore size of about 10 nm or more and less than 100 nm; iii) exposing at least part of the filtered sample to a sterol lipid; iv) obtaining at least two fractions of said sample either before or after exposing it to said sterol lipid; v) exposing the first of said fractions to a substrate carrying an immobilised mycolic acid derived antigen; vi) exposing the second of said fractions to a substrate not carrying an immobilised mycolic acid derived antigen; or storing at least part of the second of said fractions until step vii), skipping the step of exposing the second of said fractions to a substrate not carrying an immobilised mycolic acid derived antigen; vii) exposing at least part of the sample fraction exposed in step v) to a test substrate carrying an immobilised mycolic acid derived antigen and exposing at least part of the sample fraction exposed or stored in step vi) to a control substrate carrying an immobilised mycolic acid derived antigen; viii) detecting binding of antibodies to the antigen of step vii) in an end-point assay; and ix) comparing the degree or extent of binding between the test and control substrates, any observed lesser binding to the test substrate being an indicator of the presence of antibodies to the antigen in the sample that indicates active tuberculosis in the human or animal from which the samples originated. In this context it should be understood that lesser can be interpreted qualitatively and quantitatively, i.e. lesser binding may be interpreted as having less binding events as well as having weaker bindings. The advantage of this embodiment is that no sample of a healthy person is required as a reference or control sample. Only one sample from one subject is necessary in this embodiment. The sample from a human or animal suspected of having active tuberculosis of step i) of this embodiment is as described above. One exemplary way of exposing the sample to the sterol lipid is to lead the sample into a long spiral channel and pass it along the length of the channel, which is pre-coated with a sterol lipid, which is preferably cholesterol. At the end of the channel the lipid sterol exposed sample may pass a means for dividing the sample stream such as a passive valved branch point that leads the divided sample streams to the next substrate carrying an immobilised mycolic acid derived antigen of step v) and the substrate not carrying an immobilised mycolic acid derived antigen of step vi), e.g. a container comprising a substrate carrying an immobilised mycolic acid derived antigen, and a container comprising a substrate not carrying an immobilised mycolic acid derived antigen respectively. An alternative exemplary way of exposing the sample to the sterol lipid is to inject it into a container coated with cholesterol, followed by incubation therein for less than 10 minutes. Another alternative exemplary way is to expose the sample to the sterol lipid, wherein the sterol lipid is contained in a compartment of a column. In such a compartment the sterol lipid is preferably coated on beads. Whole blood/plasma/serum is rich in hundreds different kind of molecules with hydrophilic to hydrophobic properties. Therefore substrate material will be used that is inert for non-specific binding of molecules of this sample. In this context the substrate should be understood to be a support material. In the next step of this embodiment a fraction of the sterol lipid exposed sample is exposed to a substrate carrying an immobilised mycolic acid derived antigen (step v)) and another fraction of the sterol lipid exposed sample to a substrate not carrying an immobilised mycolic acid derived antigen (step vi)). To obtain at least two fractions, at a certain moment in the process the blood/plasma stream has to be divided (step iv) . For sake of convenience and to optimize reliability of the method, the stream may be divided after exposure to the sterol lipid, i.e. that the sterol lipid exposed sample is divided into at least two fractions to provide a test stream and a control stream. It is preferred that the sample is divided in two equal or substantially equal fractions, because this enables a simple comparison of both fractions without the need for correction calculations. Dividing the sample stream may be carried out by any means for dividing the sample stream such as a passive valved branch point. Alternatively the sample stream (which is filtered to plasma or serum) is divided into at least two fractions after the filtering step, but before exposure to the sterol lipid. Alternatively the blood stream is divided before filtering. This is less preferred because in this case two filters would be required. In case columns are used the stream is in this case divided before exposure to the sterol lipid in step iii) , so that steps iii) and v) can take place in the same column and step iii) and vi) can take place in another column. The substrate carrying an immobilised mycolic acid derived antigen and the substrate not carrying an immobilised mycolic acid derived antigen of steps v) and vi) are preferably of the same material. In addition, since sterol lipids, (e.g. cholesterol) and mycolic acid derivatives are both hydrophobic, the substrates for exposure in steps v) and vi) may be of the same material as the substrate that may be used to in step iii). It will be understood that suitable substrate material is inert for non-specific binding of molecules of this sample. The exposure time of the sample to the sterol lipid is preferably less than 10 minutes, such as 2 to 8, 3 to 7, 4 to 6 or about 5 minutes. One exemplary way of exposing the sample to the substrates of step v) and vi) is to lead the divided sample streams into long spiral channels, preferably implemented in micro-chips, pre-coated either with a mycolic acid derivative (step v) or without a mycolic acid derivative (step vi) and pass it along the length of these channels. An alternative exemplary way of exposing the divided lipid sterol exposed samples to a substrate coated either with a mycolic acid derivative or without a mycolic acid derivative is to inject the a first stream into a container comprising a substrate carrying an immobilised mycolic acid derived antigen (step v) , and a second sample stream into a container comprising a second substrate not carrying an immobilised mycolic acid derived antigen (step vi) and incubate the samples for less than 10 minutes, such as 2 to 8, 3 to 7, 4 to 6 or about 5 minutes. Another alternative way is that steps v) and vi) take place in a column. The stream is in this case divided before exposure to the sterol lipid in step iii), so that step iii) and v) can take place in one column and step iii) and vi) can take place in another column. Although for reasons of reproducibility and reliability of the detection method it is preferred that in step vi) said another fraction of the sterol lipid exposed sample is exposed to a substrate not carrying an immobilised mycolic acid derived antigen, it may also be possible that the step of exposing said fraction to a substrate not carrying an immobilised mycolic acid derived antigen is skipped and that said fraction is stored until further use, such as in step vii) . This may for instance take place in the same column or container wherein the incubation with the sterol lipid took place. In step vii) of this embodiment the divided sterol lipid exposed sample streams, either exposed to mycolic acid derivatives in step v(test stream) or not in step vi (control stream) are passed to detection substrates, test and control substrates respectively, carrying mycolic acid derivatives, preferably the same derivatives as in step v). The test and control substrates may be any of the detection substrate as described above for end-point methods.

Mycolic acid derived antigens

The mycolic acid derived antigens as referred to in the present application may be derived from mycobacteria selected from virulent and pathogenic mycobacteria. Preferably, the mycolic acid antigen is derived from Mycobacterium tuberculosis. Said mycolic acid derived antigen is at least one selected from the group of mycolic acid, cord factor, chemically modified mycolic acid, chemically modified cord factor, a synthetic mycolic acid derivative, a synthetic cord factor derivative.

The mycolic acid derived antigen may suitably be selected from one or more of mycolic acids obtained from natural sources, synthetically prepared mycolic acids, sulfur-containing mycolic acids, structural analogues of mycolic acids, and mycolic acid wax esters. The mycolic acid derived antigen also includes salts and/or esters of these derivatives.

Natural sources of mycolic acid derivatives include the cell walls of mycobacteria such as Mycobacterium tuberculosis include mixtures of different classes of compounds and different mycolic acid homologues, often as derivatives in which they are bonded to the wall of the cell.

It may be preferred to use synthetically prepared mycolic acids because then it can be exactly determined which and which amount of a particular derivative is used. This is advantageous for obtaining a high substrate selectivity.

Therefore in order to be able to provide a detection with high reliability and reproducible results it is preferred to use semi-synthetic or even more preferred synthetic mycolic acid derivatives which are identical or closely analogous to single compounds found in natural mixtures .

Esters of mycolic acid derivatives can also be used such as esters of alcohols (e.g. monohydric alcohols and polyhydric alcohols) and sugar esters. Sugar esters are particularly preferred. Sugar esters include esters with a monosaccharide, disaccharide or an oligosaccharide. Said saccharides may conveniently include sugar units based on hexoses and/or pentoses. Glucose esters are suitable examples of these esters. Further suitable sugar esters are trehalose esters, including trehalose monomycolates and trehalose dimycolates. For instance cord factors, which are trehalose monomycolates or trehalose dimycolates are well known examples of sugar esters that are suitable. These compounds occur in nature as complex mixtures of different classes of mycolic acids and of different homologues within each class.

Because it is difficult to establish the identity of cord factors present in natural products and to separate individual molecular species it is preferred to use semisynthetic or more preferably synthetic mycolic acid derivatives for the purposes of the invention. Further, it is known that the structure of the mycolic acid unit affects the biological activity of the cord factor. Suitable semi-synthetic derivatives include semi-synthetic cord-factors which may be prepared by attaching mycolic acids to the sugar group. These semi-synthetic factors however still contain mixtures of different homologue. Therefore particular suitable mycolic acid derivatives for use in the context of the present invention are synthetic cord factors, for example the synthetic cord factors described in WO 2010/08667, i.e. compounds of formula (M) x (S) y (M')z, wherein x is from 1 to 6, y is from 1 to 12, z is from 0 to 10, each M and each M' is independently a mycolic acid residue including a β-hydroxy acid moiety and each S is a monosaccharide unit.

Salts of mycolic acid derivatives can also be used, for instance ammonium salts, or alkali metal and alkaline earth metal salts.

Sulfur-containing mycolic acids and/or esters or salts thereof may also be used. These compounds are analogues of natural mycolic acid compounds containing sulfur .

Mycolic acid wax esters comprise a cyclopropyl group or an alkene and an internal ester group and can be isolated from natural sources or prepared synthetically.

Suitable compounds for use in the detection method of the invention are described inter alia in WO 2016/024116.

The mycolic acid antigen may be in a form selected from homogenous and heterogenous compound mixtures. The mycolic acid derived antigen may for instance be used in combination with a phospholipid such as phosphatidylcholine .

The mycolic acid antigen may be immobilised on the substrates in various ways that are known to the skilled person. Synthetic mycolic acid derived antigens may be synthesised with particular active groups that enable immobilisation to a substrate material.

For instance, for immobilisation on silica, silane coupling chemistry may be applied.

In case of a nitrocellulose substrate the mycolic acid derived antigen may be suitably immobilised as follows. Mycolic acid derived antigens may be obtained in lyophilized form and be reconstituted in a solvent mixture, for instance a chloroform: methanol: water mixture, and diluted to a concentration in the order of several nanomolars, for instance 1 nM. This dilution can then be spotted on a nitrocellulose membrane, wherein each spot is separated at a predetermined distance, e.g. 1 cm. After drying of the spots the mycolic acid derived antigens are immobilized on the membrane. Alternative immobilisation methods may for instance involve dissolving the antigens in hexane or hot PBS to form an antigen coating solution before spotting the solution on a membrane .

Examples

The following example is meant to illustrate the principle of the invention and should not be interpreted as limiting the scope of the claims. In the example a human serum sample was tested for antibodies against mycolic acid derivatives in an ELISA assay. As a detection method an ELISA method was chosen because it is less sensitive than for example Surface Plasmon Resonance or electrochemical impedance spectroscopy (EIS) or Ring Resonance (Interferometry). Therefore it can be concluded that if satisfactory results are obtained with ELISA, these will also be obtained with more sensitive detection methods .

Materials : • Sample (human plasma or serum) • 0.2 micron spinfliters (Whatman) • Mycolic Acid and Cholesterol-coupled beads • polyclonal Mycolic Acid dissolved in hexane • Polystyrene ELISA plates

• PBS

• Blocking buffer: 0.5% casein in PBS • 1-step ultra TMB-ELISA substrate solution (Thermo Scientific) • 2M sulfuric acid

• Secondary antibody: rabbit anti-human Ig HRP (DAKO)

Procedures

Coating of ELISA plates

To each well of 96-wells ELISA plates 50 μΐ (of hexane with polyclonal mycolic acid in a concentration of 3 pg/ml was added. Hexane without mycolic acid was used as control. Plates were incubated for 24 hrs at 4°C and subsequently washed two times with PBS.

Preparation of the beads

In the present example cholesterol was coupled to beads. For purpose of this example Toyopearl AF-Amino-650M beads (Tosoh Bioscience) were used. Beads were prepared in accordance with the manufacturer's instructions. As cholesterol 7-keto cholesterol was conjugated to the beads basically as described in Abdel-Magid AF et al. (Reductive Amination of Aldehydes and Ketones with Sodium

Triacetoxyborohydride. Studies on Direct and Indirect Reductive Amination Procedures. J. Org. Chem. 1996:61,-3849-38 62) by mixing 7-keto cholesterol (100 mg, 0,25 mmol) and Toyopearl slurry (2,5 ml, 0,25 mmol reactive groups) in acetonitrile (3,5 ml), followed by treating with sodium triacetoxyborohydride (80 mg, 0,375 mmol). The mixture was stirred at RT under a N2 atmosphere for 1,5 h. After that, the reaction mixture was quenched by adding aqueous saturated NaHCCt. The beads were washed thoroughly with H20, 1M NaCI, and finally H20 to remove excess ligand. After that the residual amino groups were acetylated by adding 8 ml of 0,2M sodium acetate and 4ml of acetic anhydride to the resin at 0°C for 30 minutes followed by adding another 4 ml of acetic anhydride and incubation at 25°C for 30 minutes.

Sample preparation

Serum derived from humans that were known to be suffering of tuberculosis was diluted 1:5 in blocking buffer (TB+ samples) . The TB+ samples used were derived from both smear positive and smear negative patients. Also samples were used derived from healthy humans (Control samples). Two fractions of 0,5 ml were obtained and transferred each to a 0,2 micron spin filter and centrifuged at lOOOOg. The flow-through was pooled from 1:5 to 1:20 in blocking buffer.

After this 0,2 micron spin filter samples were filtered through cut-off membrane filters. Several cut-off membrane filters were tested. As a control, the cut-off membrane filter step was skipped, so that the only filter step for controls is the 0,2 micron spin filter step. The filters used were Nanosep® filter systems of Pall Corporations and were used in accordance with the manufacturer's instructions using spin filtration at 10000 g. The filter medium of Nanosep® is Omega™, which is low protein-binding, modified polyethersulfone. Filters were used with Molecular Weight Cut Off (MWCO) of 10 kDa, 100 KDa, 300 KDa and 1000 KDa. This corresponds to membrane nominal pore sizes of ~1 nm, 10 nm, 35 nm and 100 nm respectively.

In case of exposure of a sample to beads with cholesterol, after the cut-off membrane filter treatment 250 μΐ of the flow-through was added to beads which were either coated with cholesterol, polyclonal mycolic acids or blocking agent. Beads were agitated at room temperature. After incubation the beads were spun down fast and the supernatant was used as a sample for antibody detection.

The total pre-treatment of the samples in this particular set up may take place in less than 10 minutes, after which analysis can directly follow. ELISA-procedure

To block aspecific binding of antibodies with the mycolic acids in the wells of the ELISA plates, 300 μΐ of blocking buffer was added to each well and incubated for 1 hour. Subsequently, the blocking buffer was replaced with 45 μΐ of sample. Blocking buffer was used as a negative control. After incubation, the plates were washed three times with PBS. The washing PBS buffer was replaced with HRP-conjugated secondary antibody in blocking buffer. After incubation, the plate was washed again three times with PBS. Subsequently 50 μΐ per well of TMB-ELISA substrate solution was added to the wells and incubated for 15-30 min at room temperature. The reaction was stopped by adding 50 μΐ per well of 2 M sulphuric acid. Absorbance at 450 nm was measured to quantify the binding of antibodies to the mycolic acid substrate of the ELISA plate .

Tests

Influence of cut-off membrane filter and cholesterol exposure

To test the effect of using a cut-off membrane filter treatment samples 4 sets of samples were subjected to different treatment. Except for differences in cholesterol exposure and the use of a cut-off membrane filter treatment all other conditions were the same for all treatments and were as described above.

In one treatment samples were not exposed to cholesterol and not filtered with a 300 MWCO filter before ELISA.

In another treatment samples were not exposed to cholesterol but were filtered with a 300 MWCO filter before ELISA.

In another treatment samples were exposed to cholesterol but were not filtered with a 300 MWCO filter before ELISA.

In another treatment samples were exposed to cholesterol but were filtered with a 300 MWCO filter before ELISA.

For each treatment 6 samples from different individuals (in case of TB+ samples SMEAR + and SMEAR -) were subjected to treatment as described above. For each treatment the samples were from the same 6 individuals. Two tests were performed under the same conditions but at different times. The average ELISA signal obtained from 6 samples subjected to the same treatment was determined. The ELISA signal is shown in Table I below.

Table 1 ELISA results

The ELISA signal of the control samples can be considered as background signal as these samples are derived from healthy humans that should not test positive for tuberculosis. As expected the TB+ samples give a higher signal than the control samples derived from healthy humans, thus indicating the presence of antibodies against mycolic acid (derivatives).

In test 1, for samples that were not exposed to cholesterol and not filtered with a 300 MWCO filter before ELISA, the percentage of background signal was 0,825/2,249 = 37%. In test 1, for samples that were not exposed to cholesterol and which were filtered before ELISA, the percentage of background signal was thus 0,044/ 0,301= 15%. In test 2, for samples that were not exposed to cholesterol and not filtered with a 300 MWCO filter before ELISA, the percentage of background signal was 0,690/2,071 = 33%. In test 2, for samples that were not exposed to cholesterol and which were filtered before ELISA, the percentage of background signal was thus 0,0476/ 0,313= 15%. From this it follows filtering a sample from a human or animal suspected of having active tuberculosis using a cut-off membrane filter with a nominal pore size of 300 MWCO, which corresponds to a membrane nominal pore size of 35 nm before exposing the sample to a substrate carrying an immobilised mycolic acid derived antigen and detecting binding of antibodies to the antigen results in a reduction of background signal of more than 50%.

It was further tested whether the background signal could be further decreased by exposing the sample to cholesterol after filtrating it with said 300 MWCO filter. For this purpose samples were exposed to beads with cholesterol immobilised thereon as described above.

In test 1, for samples that were exposed to cholesterol but not filtered with a 300 MWCO filter before ELISA, the percentage of background signal was 0,675/2,269 = 30%. In test 1, for samples that were exposed to cholesterol and which were filtered before ELISA, the percentage of background signal was thus 0,021/ 0,276= 8 %. In test 2, for samples that were exposed to cholesterol but not filtered with a 300 MWCO filter before ELISA, the percentage of background signal was 0,630/2,529 = 25%. In test 2, for samples that were exposed to cholesterol and which were filtered before ELISA, the percentage of background signal was thus 0,0296/ 0,276= 11%.

The results of these tests are summarized in table 2 below:

Table 2 summary of results of test 1 and 2: background signal

From this it follows that filtering a sample from a human or animal suspected of having active tuberculosis using a cut-off membrane filter with a nominal pore size of 300 MWCO, which corresponds to a membrane nominal pore size of 35 nm in combination with exposing the sample to cholesterol before exposing the sample to a substrate carrying an immobilised mycolic acid derived antigen and detecting binding of antibodies to the antigen results in an even stronger reduction of background signal.

Influence of pore size

To test the influence of pore size on the reduction of background signal cut-off membrane filters with different nominal pore sizes were tested.

Prior to any cut-off membrane filter treatment samples were filtered with 0,2 micron spin filter. Cut-off filters were used with Molecular Weight Cut Off (MWCO) of 10 kDa, 100 KDa, 300 KDa and 1000 KDa. This corresponds to membrane nominal pore sizes of 5 nm, 10 nm, 35 nm and 100 nm respectively. As a control, the cut-off membrane filter step was skipped. Again for each treatment 6 samples from different individuals (in case of TB+ samples SMEAR + and SMEAR -) were subjected to treatment as described above. For each treatment the samples were from the same 6 individuals. The average ELISA signal obtained from 6 samples subjected to the same treatment was determined and standardised to the signal obtained from the treatment in which only a 0,2 micron filter was used. The ELISA signal of the samples in which only a 0,2 micron filter was used was set at 100%. The average ELISA signal is shown in Table 3 below. Standard deviations are also indicated.

Table 3: effect of pore size on the reduction of background signal

The data in table 3 above are represented in the diagram in Fig.l. When samples were not subjected to cutoff filter treatment (see 0.2 micron bars in Fig.l) both ELISA signals for the TB+ and control samples were set at 100%. When samples were subjected to cut-off filter with 1000 kDa MWCO (corresponding to a nominal pore size of 100 nm) it can be seen that the ELISA signal of both the TB+ and control samples are a bit lower than the signals of the samples that were not subjected to cut-off filter treatment (see 1000 kD bars in Fig.l) . Apparently some of the material in the sample that would normally bind to the immobilised mycolic acid on the ELISA substrate has been scavenged away by the 1000 kDa MWCO treatment. As the relative signals of the TB+ and control samples compared to the signals of the only 0,2 micron filter treatment do not differ significantly it appears that the background

signal has not been specifically reduced. However, if a cut-off filter treatment with a cut-off of less than 1000 kDa is used, the background signal derived from aspecific binding to the immobilised mycolic acid on the ELISA substrate becomes actually lower. This is clearly shown in the 300 kDa and 100 kDa bars (corresponding to cut-off filter treatment using a 300 kDa MWCO and 100 kDa MWCO, i.e. 35 and 10 nm nominal pore size) . Here it can be seen that in the control samples the ELISA signal relative to the control samples only treated with a 0,2 micron filter is significantly more reduced than the ELISA signal of the TB+ samples relative to the TB+ samples only treated with a 0,2 micron filter. This shows that materials in the sample that would bind to the immobilised mycolic acid on the ELISA substrate and which are not specific for tuberculosis have been scavenged away by the cut-off filter treatment. In line with this, the ELISA signal derived from mycolic acid specific antibodies in the sample from tuberculosis patients becomes more pronounced, because the background signal caused by unspecific binding to the test substrate has been reduced. From this it follows that filtering a sample from a human or animal suspected of having active tuberculosis using a cut-off membrane filter with a nominal pore size of 10 nm or more and less than 100 nm, before exposing the sample to a substrate carrying an immobilised mycolic acid derived antigen and detecting binding of antibodies to the antigen results in strong reduction of background signal. This way the risk of testing false positive for tuberculosis is reduced. Accordingly the method of the invention herewith provides a method of detecting a marker for active tuberculosis which is more sensitive than the methods known in the art.

Claims

A method for detecting an active tuberculosis marker, comprising the steps of: (I) providing a sample from a human or animal suspected of having active tuberculosis; (II) filtering the sample when using a bounding membrane filter with a nominal pore size of 10 nm or more and less than 100 nm; (III) exposing at least a portion of the filtered sample to a substrate that carries an immobilized mycolic acid derived antigen; and (IV) detecting antibody binding to the mycolic acid-derived antigen.

The method of claim 1, wherein detecting antibody binding to the antigen takes place in real time.

The method of claim 1, wherein detecting antibody binding to the antigen is by means of an endpoint test.

The method of any one of claims 1 to 3, wherein at least a portion of the sample is exposed to a sterol lipid prior to its exposure to the substrate carrying an immobilized mycolic acid derived antigen.

The method of any one of claims 1 to 4, wherein at least a portion of the sample is exposed to the sterol lipid after filtering it with the boundary membrane filter, and prior to exposing it to the substrate containing an immobilized mycolic acid carries antigen.

The method of claim 5, comprising i) providing a sample from a human or animal suspected of having active tuberculosis; ii) filtering the sample when using a bounding membrane filter with a nominal pore size of 10 nm or more and less than 100 nm; iii) exposing the sample to a sterol lipid; iv) obtaining at least two fractions of the sample before or after exposing it to the sterol lipid; v) exposing the first of the fractions to a substrate that carries an immobilized mycolic acid derived antigen; vi) exposing the second of the fractions to a substrate that does not carry an immobilized mycolic acid derived antigen; vii) exposing the sample fraction exposed in step v) to a test substrate that carries an immobilized mycolic acid-derived antigen and exposing the sample fraction exposed in step vi) to a control substrate that carries an immobilized mycolic acid-derived antigen; viii) detecting the binding of antibodies to the antigen of step vii) in real time; and ix) comparing the degree or extent of binding between the test and control substrates, any observed less binding to the test substrate being an indicator of the presence of antibodies to the antigen in the sample that are active tuberculosis in humans or the human animal from which the samples originate.

A method according to claim 2 or 4-6, wherein detection is carried out using surface plasmon resonance or electrochemical impedance spectroscopy, isothermal titration calorimetry, biolayer interferometry, optical grids, photonic crystal, acoustic resonance profiling, quartz crystal microbalances.

A method for detecting an active tuberculosis marker according to claim 5, comprising i) providing a sample from a human or animal suspected of having active tuberculosis; ii) filtering the sample when using a bounding membrane filter with a nominal pore size of 10 nm or more and less than 100 nm; iii) exposing at least a portion of the filtered sample to a sterol lipid; iv) obtaining at least two fractions of the sample before or after exposing it to the sterol lipid; v) exposing the first of the fractions to a substrate that carries an immobilized mycolic acid derived antigen; vi) exposing the second of the fractions to a substrate that does not carry an immobilized mycolic acid derived antigen; or storing at least a portion of the second of the fractions to step vii), wherein the step of exposing the second of the fractions to a substrate that does not carry an immobilized mycolic acid derived antigen is skipped; vii) exposing at least a portion of the sample fraction exposed in step v) to a test substrate that carries an immobilized mycolic acid derived antigen and exposing at least a portion of the sample fraction exposed or saved in step vi) to a control substrate that carries an immobilized antigen derived from mycolic acid; viii) detecting the binding of antibodies to the antigen of step vii) in an endpoint test; and ix) comparing the degree or extent of binding between the test and control substrates, any observed less binding to the test substrate being an indicator of the presence of antibodies to the antigen in the sample that are active tuberculosis in humans or indicates the animal from which the samples originated.

The method according to claims 3-5 or 8, wherein the endpoint test is selected from the group consisting of enzyme-linked immunosorbent test (ELISA), Western blot, radioactive labeling test, photospectrometric test, immunofluorescence, immunoprecipitation, immunocytochemistry, immunohistochemistry, electrochemical impedance spectroscopy.

The method of claim 9, wherein the endpoint test is an immunogold filtration test.

The method of claim 10, wherein the immune gold test is a dot immuno gold test (DIGFA).

A method according to any one of claims 3-5 or 8-11, wherein detection takes place visually, preferably with the naked eye.

A method according to any one of claims 3-5 or 8-12, wherein detection results in a visual signal which is analyzed by means of a mobile application designed to compare the binding signal from different samples or sample fractions and which indicates or the human or animal from which the sample originates has active tuberculosis.

A method for pretreating a sample from a human or animal suspected of having active tuberculosis, comprising the steps of: I) providing a sample from a human or animal suspected of having active tuberculosis; (II) filtering the sample when using a bounding membrane filter with a nominal pore size of 10 nm or more and less than 100 nm.

The method of claim 14, wherein at least a portion of the sample is exposed to a sterol lipid before or after filtering it with the boundary membrane filter.

The method of any one of the preceding claims, wherein the nominal pore size is between 25 and 75 nm.

The method of the preceding claim, wherein the nominal pore size is between 25 and 45 nm.

The method of the preceding claim, wherein the nominal pore size is approximately 35 nm.

The method of any one of the preceding claims, wherein the membrane of the boundary membrane filter is a hydrophilic membrane.

The method of any one of the preceding claims, wherein the membrane of the boundary membrane filter is made of polyether sulfone.

A method according to any one of the preceding claims, wherein the filtering step takes place by means of ultracentrifugation.

A method according to any of claims 1-13 or 16-21 if dependent on any of claims 1-13, wherein the mycolic acid-derived antigen comprises at least one selected from the group of mycolic acid, cord factor, chemically modified mycolic acid, chemically modified cord factor, a synthetic mycolic acid derivative, a synthetic cord factor derivative.

The method of any one of claims 4-13 and 15-22, wherein the sterol lipid is immobilized on a surface.

The method of claim 23, wherein the sterol lipid is immobilized on a substrate comprising the sterol lipid and a compound that binds to cholesterol in the sample, such as pectin, amphothericin B or β-cyclodextrin.

The method of claim 24, wherein the sterol lipid is immobilized on a substrate comprising the sterol lipid and phosphatidylcholine.

The method of any one of claims 4-13 and 15-25, wherein the sterol lipid is cholesterol.